In [ ]:

# /// script
# requires-python = ">=3.12"
# dependencies = [
#     "geopandas",
#     "h3",
#     "lonboard",
#     "matplotlib",
#     "palettable",
#     "pyarrow",
#     "sidecar",
# ]
# ///

# /// script # requires-python = ">=3.12" # dependencies = [ # "geopandas", # "h3", # "lonboard", # "matplotlib", # "palettable", # "pyarrow", # "sidecar", # ] # ///

Kontur Global Population Data¶

This example will use the H3Layer with data from Kontur's global population dataset.

In [ ]:

import warnings

import geopandas as gpd
from matplotlib.colors import LogNorm
from palettable.colorbrewer.diverging import BrBG_10
from sidecar import Sidecar

from lonboard import H3HexagonLayer, Map
from lonboard.colormap import apply_continuous_cmap

import warnings import geopandas as gpd from matplotlib.colors import LogNorm from palettable.colorbrewer.diverging import BrBG_10 from sidecar import Sidecar from lonboard import H3HexagonLayer, Map from lonboard.colormap import apply_continuous_cmap

We'll load this data directly

In [ ]:

url = "https://geodata-eu-central-1-kontur-public.s3.eu-central-1.amazonaws.com/kontur_datasets/kontur_population_20231101_r4.gpkg.gz"

url = "https://geodata-eu-central-1-kontur-public.s3.eu-central-1.amazonaws.com/kontur_datasets/kontur_population_20231101_r4.gpkg.gz"

In [ ]:

# Wrap with GDAL's virtual file prefixes
vsipath = f"/vsigzip//vsicurl/{url}"

# Wrap with GDAL's virtual file prefixes vsipath = f"/vsigzip//vsicurl/{url}"

We use warnings.catch_warnings to avoid the logging of GDAL not being able to find auxiliary files in the bucket.

In [ ]:

with warnings.catch_warnings():
    warnings.filterwarnings(
        "ignore",
        message="HTTP response code",
        category=RuntimeWarning,
    )
    gdf = gpd.read_file(vsipath, use_arrow=True)

with warnings.catch_warnings(): warnings.filterwarnings( "ignore", message="HTTP response code", category=RuntimeWarning, ) gdf = gpd.read_file(vsipath, use_arrow=True)

We extract only the attribute columns in this case so that we render the h3 data directly, instead of rendering the polygon data.

This now gives us a plain pandas DataFrame, we no longer have a geometry in df.

In [ ]:

df = gdf[["h3", "population"]]

df = gdf[["h3", "population"]]

We create a color palette where the colormap is applied to the population values logarithmically.

In [ ]:

pop = df["population"]
min_bound = pop.min()
max_bound = pop.max()
normalizer = LogNorm(min_bound, max_bound, clip=True)
normalized = normalizer(pop)

pop = df["population"] min_bound = pop.min() max_bound = pop.max() normalizer = LogNorm(min_bound, max_bound, clip=True) normalized = normalizer(pop)

Create the map window for display:

In [ ]:

sidecar = Sidecar(anchor="split-right")

sidecar = Sidecar(anchor="split-right")

Finally we generate the layer and pass it into a map.

In [ ]:

layer = H3HexagonLayer.from_pandas(
    df,
    get_hexagon=df["h3"],
    get_fill_color=apply_continuous_cmap(normalized, BrBG_10, alpha=0.7),
)
m = Map(layer)
with sidecar:
    display(m)

layer = H3HexagonLayer.from_pandas( df, get_hexagon=df["h3"], get_fill_color=apply_continuous_cmap(normalized, BrBG_10, alpha=0.7), ) m = Map(layer) with sidecar: display(m)

Due to an upstream bug, you may temporarily need to move the map slightly for the data to show.